You do not have to pick a number. Paira builds each session by splitting it three ways — roughly half review, about a third new words, the rest a refresh of things you have seen — so the number of new words you meet falls out of how well you are doing rather than out of a target you set in advance. On a good day you meet more new words. On a day when you keep failing, review crowds them out, which is the correct response.
Why a fixed daily target misleads
The advice you find for this question is nearly always a bare number: ten a day, twenty a day, fifty if you are serious. The trouble is that a new word is not a one-off cost. Every word you meet today books future appointments with you — tomorrow, then in two days, then four, seven and fourteen.
So "twenty new words a day" is really a promise to handle twenty new words plus everything the previous fortnight of new words is due to send back. People who commit to a high number usually meet it for about a week, then quietly stop, because the review debt arrives all at once and the session stops feeling like progress.
The review intervals that turn today's new words into next week's queue
How a session is actually built
Paira fills each round from three buckets and interleaves them:
- Review, around half the round — words that are due or weak, hardest first: lowest mastery, then most overdue
- New, around a third — words you have never been shown, in the order they appear in the set
- Refresh, the remaining sixth — words you have seen and are not struggling with, least recently seen first
In a twenty-word session that is roughly ten review, seven new and three refresh. Those are proportions rather than quotas, and a short bucket does not leave a gap: if there are only four words due, the rest of the review share is backfilled from new, and if new runs out too, from refresh. A quiet review day therefore turns into a bigger intake of new words automatically.
The number that actually caps you
It is not your appetite for new words. It is how many words you are currently failing. Every wrong answer drops that word's mastery and makes it due immediately, so it lands back in the review bucket of the very next session. Twelve words you keep missing will fill the review half of your session on their own, and the new-word share shrinks accordingly.
That is a feedback loop doing its job. When you are struggling, taking on more new vocabulary makes the struggle worse, and the split is what stops you from doing it.
So what should you do instead of picking a number?
Pick a frequency, not a quantity. One session a day is a decision you can keep; twenty new words a day is a decision that gets harder every day for two weeks and then breaks. The interval schedule assumes you come back roughly daily, and it degrades gracefully if you miss a day, because overdue words simply sort to the front.
If you want to move faster, do a second session rather than raising a target. The split applies again, so the extra session spends about half its effort on review as well, and you get the additional new words without leaving the debt unpaid.